This is a plain-English summary of a fund's half-year report. The fund made 36.6% over six months, beating its benchmark (a regional stock index) which returned 17.8%. It gained from big bets on Korean chip makers, copper and gold miners, and Chinese tech stocks. For everyday investors, the key point is that the fund now leans heavily on China and Korea and uses more borrowed money, which can boost gains but also increases losses. It also bought back its own shares to support the stock price. A shareholder vote in November 2026 will decide whether the fund continues. Worth reading if you want a clear example of how a focused, aggressive investment trust works.
Pacific Horizon's interim report shows that for the six months ended 31 January 2026, the company's total return on net asset value (NAV) per share reached 36.6%, far exceeding the 17.8% of the MSCI All Country Asia ex Japan Index (in sterling); the total return on the share price was 38.7%. Shareho
NAV total return for the reporting period was 36.6%, well ahead of the 17.8% for the sterling-based benchmark; the share price total return was 38.7%, and the discount narrowed from 9.5% to 8.1%.
| Metric | Six months to 2026-01-31 | Year to 2025-07-31 |
|---|---|---|
| NAV total return | 36.6% | 8.3% |
| Share price total return | 38.7% | 6.4% |
| MSCI All Country Asia ex Japan (in sterling) | 17.8% | 17.1% |
| MSCI index price change (original-table basis) | 16.9% | — |
Note: The column order of the summary table in the incoming text appears to be reversed; the table above has been reconstructed to compare 2026-01-31 against 2025-07-31, based on the Chairman's statement and the period's high/low price range. After reconstruction: shareholders' funds £809.9m (2026-01-31) versus £613.4m (2025-07-31); NAV 716.13p → 976.21p; share price 648.00p → 897.00p.
Excess return in the period came almost entirely from stock selection, concentrated in three areas: North Asian semiconductors, copper/gold, and Chinese technology; the report provides only qualitative attribution and does not quantify individual-stock contributions.
The report names four positive contributors: SK Square, Samsung Electronics, MMG, and Zijin Mining Group; it also notes that "certain Chinese technology stocks (unnamed)" were positive contributors as well.
| Holding | Contribution/Detraction | Direction (as per original) | One-line attribution |
|---|---|---|---|
| SK Square | Positive (not quantified) | Not stated | North Asian semiconductor holding, lifted by broadening AI adoption |
| Samsung Electronics | Positive (not quantified) | Not stated | Semiconductor leader and its ecosystem beneficiaries |
| MMG | Positive (not quantified) | Not stated | Higher copper prices plus strong operational delivery |
| Zijin Mining Group | Positive (not quantified) | Not stated | Higher copper and gold prices drove substantial share price gains |
| Chinese technology stocks (unnamed) | Positive (not quantified) | Not stated | Technology leadership spreading from AI pioneers to mass-adoption enablers |
Stance flagged as [cautiously optimistic]: the long-term structural logic is clear, but near-term risks centre on technology valuations, index concentration, and macro/geopolitical uncertainty; the manager explicitly stresses "selective investing".
The report discloses no buy/sell details for any individual stock; the visible portfolio-level changes are higher leverage, a lower active share, and large-scale buybacks that shrank the share count.
Revenue earnings turned from positive to negative during the period, buyback intensity remained high, and the Board noted that a quinquennial continuation vote will be held in November 2026.
From the position changes disclosed in the subsequent section, the core move of the half-year was a decisive "great shift": China was substantially increased, while India was systematically reduced. China's weight rose to 39.3% (an overweight of 8.9%), making it the largest relative position; India was halved from 14.2% to 5.4%, turning it into the largest underweight (-9.9%). This extreme regional tilt is not a simple valuation comparison but a re-rating of earnings-growth certainty over the next three to five years:
| Region | Weight six months ago | Current weight | Versus benchmark |
|---|---|---|---|
| China | ~30% (estimated) | 39.3% | +8.9% |
| India | 14.2% | 5.4% | -9.9% |
India is not a case of broken fundamentals; rather, valuations are "demanding", leaving insufficient odds. China, by contrast, is experiencing a threefold convergence of "manufacturing upgrade + AI hardware demand + commodity supply chains". Notably, of the roughly 1200bps of net additions to China, about 375bps came from Alibaba, with the remainder spread across AI, new-energy, and mining companies — indicating that this is an increased bet on the combined "new economy + resource security" play.
NAV total return of 36.6% versus a benchmark of 17.8% in the period implies an excess return of 18.8%. Attribution shows that SK Square alone contributed 640bps (share price up 248%), Samsung Electronics contributed 220bps (up 110%), and the materials sector contributed 620bps. This means the three top-ranked positions contributed nearly all of the excess return. This approach of "concentrated bets on critical nodes" depends heavily on accurately identifying bottleneck positions in the value chain:
| Position/Sector | Attribution contribution | Key driver |
|---|---|---|
| SK Square | 640 bps | HBM demand + supply contraction; share price +248% |
| Samsung Electronics | 220 bps | Memory price increases + foundry diversification expectations; share price +110% |
| Materials sector overall | 620 bps | Stronger copper prices, record gold prices, record-high miner output |
Active share fell from 82% to 68%. On the surface, the portfolio has become more "passive", but the manager explicitly attributes this to "holding the index's largest weights out of conviction, not compromise". In fact, when TSMC, Samsung, Tencent, Alibaba, and SK Hynix already account for roughly one-third of the index, holding these names heavily itself means tracking the index on a large scale. The key point, however, is that these five companies happen to represent the strongest competitiveness in AI computing power, memory, applications, and platform economics. The decline in active share is the natural result of high-conviction bets — only these bets are concentrated in the most crowded but most soundly profitable tracks.
From a risk perspective, leverage rising from 5% to 7%, together with buybacks of 3.1% of share capital, shows that the manager is more than confident, but it also amplifies NAV volatility. Should the semiconductor cycle reverse, the concentrated holdings under double leverage would suffer significant drawdowns.
The significance of the new copper and gold miner positions lies not only in earnings elasticity but in hedging the portfolio's "technology-stock beta". The copper logic (a decade of underinvestment + mine development cycles exceeding 10 years + demand from new energy and AI data centres) and the gold-miner logic (falling real rates + central bank gold purchases + geopolitical risk) reinforce each other. In particular, direct exposure to gold equities is about 5%, which, together with ETF-style holdings, can provide a partial buffer when technology stocks correct. This twin-engine "technology + resources" drive gives the portfolio a measure of balance between "growth" and "defence".
What is worth noting is not the fact of the increase but the selective additions and subtractions: exits from KE Holdings, JD.com, and Supor; reductions in Sanhua and Tsugami; purchases of Alibaba, MiniMax, and Knowledge Atlas; an increase in CATL; and a new position in Wasion Holdings. This clearly shows the manager shifting from "zero-sum competition in the consumer internet" to "AI infrastructure + high-end manufacturing + new-energy hardware". In particular, the newly bought MiniMax and Knowledge Atlas are both AI-related IPOs, which, together with the 375bps heavyweight Alibaba position, reflect long-term confidence in the commercialisation of AI applications and cloud services in China. Avoiding Xiaomi and BYD, by contrast, reflects caution about the smart EV industry's "competitive intensity + overcapacity" — and this avoidance is itself a source of excess return.
Consumption and property remain the "sore spots": SEA, Luckin, PDD, Embassy Developments, and Khang Dien all detracted from relative performance. Vietnam's property weakness has persisted for several years, and the manager maintains its "turning point is near" judgment, but the time cost cannot be ignored. In addition, the exit from Indian property, retaining only Embassy, indicates wariness of the policy cycle. These detractors stand in sharp contrast to the excess contributions from semiconductors and commodities — the portfolio's alpha is highly dependent on cyclical positioning, and if AI capex or commodity prices mean-revert, the give-back in relative returns would be rapid.
The portfolio has evolved into a "northern-lights engine": TSMC, SK Square, and Samsung form the computing-power core, China's AI + new energy form the growth wing, and copper and gold resources form the hedging wheel, earning excess returns through extreme regional and sector tilts. Although the active share has fallen and leverage has risen, it is precisely because of extreme confidence in structural trends that the manager dares to place heavy bets on index heavyweight stocks. What really needs continuous tracking is: the sustainability of the HBM cycle, whether the copper and gold supply-demand gaps materialise, and the pace of AI commercialisation in China. If the three logics continue to resonate, the 36.6% half-year return may be only the prelude to a longer cycle.
In the report's empirical section, one can observe how the investment themes described above have been translated into the holdings structure. This section focuses on the linkages among the valuation governance mechanism, portfolio changes, and financial performance.
The valuation process disclosed in the report is completely different from the public perception of private-asset valuation as "fund managers pricing off the top of their heads"; it reflects a fairly mature institutional design:
| Layer | Role | Responsibility |
|---|---|---|
| Investment team | Interested party | Receives valuation results only; no participation in pricing |
| Valuation committee | Internal independent body | Members drawn from different functions such as operations and risk control; decides valuations by vote |
| S&P Global | External third party | Issues valuation reports, providing a benchmark reference |
The key detail is that the investment team is notified only after the valuation has already been applied, and the valuation committee tracks market benchmarks weekly and adjusts in line with S&P Global's latest reports. This "isolation + external support" model effectively avoids conflicts of interest — because position valuations directly affect the fund managers' performance.
The three-year rolling revaluation mechanism (one-third of holdings revalued each month) means that, in normal market conditions, each private investment is independently assessed an average of 4 times a year. Combined with the Board's semi-annual review and the external scrutiny of the annual audit, the valuation frequency of private assets is in fact close to the information-disclosure density of most listed companies.
Total assets at the period end were £869,883, up approximately 33% from the period start (£651,771). Net capital gains over the same period were £218,198, showing that nearly all of the growth came from investment appreciation rather than capital injection (share capital remained at 9,208, with no new issuance).
Comparing the distribution as at July 2025 with that as at January 2026, the asset allocation underwent a notable regional rotation, the most striking feature being that India's weight was nearly halved:
| Region | January 2026 | July 2025 | Change (p.p.) |
|---|---|---|---|
| China | 29.8% | 30.3% | -0.5 |
| South Korea | 21.0% | 13.6% | +7.4 |
| Taiwan | 17.4% | 16.3% | +1.1 |
| China A-shares | 9.5% | 6.1% | +3.4 |
| Vietnam | 7.7% | 9.1% | -1.4 |
| India | 5.4% | 14.2% | -8.8 |
| Hong Kong | 3.7% | 1.8% | +1.9 |
| Singapore | 2.1% | 4.2% | -2.1 |
| Indonesia | 1.1% | 1.7% | -0.6 |
| Thailand | 1.1% | 1.1% | 0 |
| Kazakhstan | 0.8% | 1.4% | -0.6 |
| Net current assets | 0.4% | 0.2% | +0.2 |
India slipped from the second-largest position region to fifth, while South Korea jumped to become the second-largest region. Looking at the holdings list: in Korea, new additions/increases were concentrated in the semiconductor value chain (Samsung, SK Square, EO Technics); in India, although names such as Delhivery, PolicyBazaar, and Bajaj Finserv were retained, weights were cut across the board. This is not merely a passive decline but the result of an active rotation toward "hardware + resources".
The changes at the industry level are even more dramatic than those at the geographic level:
| Industry | January 2026 | July 2025 | Change (p.p.) |
|---|---|---|---|
| Information technology | 37.1% | 29.0% | +8.1 |
| Industrials | 15.0% | 13.5% | +1.5 |
| Materials | 14.2% | 6.5% | +7.7 |
| Consumer discretionary | 12.4% | 14.4% | -2.0 |
| Communication services | 9.7% | 15.8% | -6.1 |
| Financials | 9.0% | 13.2% | -4.2 |
| Real estate | 1.8% | 5.6% | -3.8 |
| Energy | 0.4% | 1.8% | -1.4 |
| Net current assets | 0.4% | 0.2% | +0.2 |
Information technology plus materials grew by nearly 16 percentage points combined, while communication services, financials, and real estate together shrank by about 14 percentage points over the same period. This "hardwarisation" turn aligns with the triple narrative of AI computing demand, global supply-chain reconfiguration, and a strong precious-metals supercycle. From the holdings:
Notably, the sharp decline in communication services is consistent with the contraction in the weights of leading internet holdings such as Tencent (5.7% weight) and Alibaba (3.0% combined) — although they remain the portfolio's foundation, the direction of marginal additions has clearly shifted toward hard tech and resources.
The income statement reveals a typical structural dilemma of investment trusts — when portfolio size grows rapidly, costs grow faster than revenue:
| Item | 2026 H1 | 2025 H1 | Change |
|---|---|---|---|
| Dividend income | 3,275 | 3,626 | -9.7% |
| Investment management fee | 2,270 | 1,831 | +24.0% |
| Borrowing costs | 1,297 | 482 | +169.1% |
| Other administrative expenses | 425 | 406 | +4.7% |
| Net revenue return | -1,025 | +467 | Turned to a loss |
The decline on the income side is likely because the portfolio's shift toward semiconductor equipment names (low dividend yields) and mining names (cyclically volatile earnings) compressed current dividend income; the rise on the cost side stems from the larger management-fee base brought by higher asset values, as well as higher year-end borrowing leverage.
Borrowing costs rose by nearly 1.7 times year on year. Estimated from the balance sheet, current liabilities increased from 38,845 to 66,606, an increase of about 27,761. This was most likely used to seize market opportunities by adding to Korean technology stocks and A-share assets — an arbitrage of "low-cost financing + high-certainty assets".
Capital returns of 218,485 were more than 200 times the revenue return over the same period, making the portfolio's net return almost entirely determined by changes in market value. Capital return per share was 260.43p; relative to the period-start share price, the half-year capital return rate was roughly in the 30%-35% range (estimated on net assets of 651,771 / per-share par value, the period-start NAV was approximately 669p per share, implying a capital return rate of about 39%). This is precisely the expected shape of a high-conviction growth trust — concentrated bursts of gains, with a thin dividend safety cushion.
Aggregating the investment list gives the portfolio's concentration metrics:
| Rank range | % of total investments |
|---|---|
| Top 5 holdings | 39.7% |
| Top 10 holdings | ~55% |
| Top 20 holdings | ~71% |
| Top 50 holdings | ~95% |
The largest single holding, TSMC, accounts for 12.9% of total invested assets, Samsung 9.7%, and SK Square 8.1% — the three Korean semiconductor value-chain names alone together account for nearly 30% of the portfolio. This concentration means the trust is essentially betting on the beta of the "East Asian computing power — resources" dual main line, and its volatility is inherently higher than the industry average.
Private company investments fell from 6.1% to 4.6% of total assets, but their absolute size was almost unchanged (from approximately 39,800 to approximately 40,000). This suggests that private asset prices remained relatively stable, and the appreciation of listed stocks in the portfolio diluted their weight.
Private holdings span AI (MiniMax), content (ByteDance, Dailyhunt), biomedicine (Chime, Eden), and e-commerce logistics (Meesho) — overall still positioned in high-risk, high-elasticity frontier tracks, but the size constraint (the largest private holding, ByteDance, accounts for only 3.3%) prevents them from having a decisive impact on the portfolio.
The portfolio's character in this period can be summarised as "twin-engine drive, income–expenditure imbalance": on the asset side, a rapid switch toward computing-power hardware and mining physicals, with Korea and A-shares becoming the largest regions for additions; on the liability side, leverage was used to amplify return elasticity. The valuation governance mechanism ensures rigorous pricing of private assets, while the concentrated burst of portfolio returns confirms that, within this system, the investment team is genuinely focused on selecting long-term growth points rather than short-term valuation games.
The statement of changes in shareholders' funds and the cash flow statement reveal the company's capital strategy over the half-year, whose core features can be summarised as "borrowing to expand, buybacks to shrink capital, dividends taking a back seat".
1. Share Buybacks and Per-Share Metrics
During the period, the company bought back 2,686,152 ordinary shares (85,648,427 → 82,962,275) at a cost of £19.674m (the cash flow statement shows actual payments of £20.94m; the difference may include settlement costs). The buyback reduced issued share capital by approximately 3.14%, directly lifting net assets per share: NAV rose from 716.13p to 976.21p, an increase of 36.3%, of which approximately 1.5 percentage points came from the buyback contribution (calculated on the period-start share capital; had the share count been unchanged, NAV would have been approximately 961.5p). The buyback was funded not by operating cash flow but by bank borrowings — indicating that management believes the current share price is undervalued and is willing to add leverage to buy its own shares.
2. Leverage Levels and Financing Costs
During the period, bank borrowings increased by a net £24.22m (£60.00m drawn, £35.78m repaid), while the net financing inflow in the cash flow statement was only £2.02m, because £22.20m was spent on buybacks and dividends. Borrowing costs rose markedly:
| Item | 2026 H1 (£'000) | 2025 H1 (£'000) | Change |
|---|---|---|---|
| Loan interest | 1,277 | 300 | +325.7% |
| Non-utilisation fees | 20 | 182 | -89.0% |
| Total | 1,297 | 482 | +169.1% |
Interest expense rose sharply while non-utilisation fees fell, indicating that the amount of the facility actually drawn was far higher than in the same period last year and that leverage utilisation increased. It should be noted that borrowing costs are trivial next to capital returns of £218m, but if the market corrects, the interest burden would erode income (the revenue return is already negative).
3. Cash Flow and Earnings Quality
Despite net returns as high as £217m, operating cash flow was a net outflow of £3.988m. This is not operational deterioration; rather, accounting earnings mainly stem from unrealised investment appreciation (net investment gains of £218m), and the dividends actually received in cash were only £3.43m (£3.743m in the same period last year, -8.4% year on year). The decline in dividend income reflects the portfolio's tilt toward low-dividend growth stocks, or reduced payouts by some companies. Investing activities recorded a net inflow of £5.409m, showing that sales exceeded purchases during the period's rebalancing; however, the scale of acquisitions (£156m) and disposals (£161m) indicates high turnover, and transaction costs warrant attention (note: not itemised separately in the disclosures).
1. Tiered Management Fee Effect
Management fees are charged at 0.75% (on the first £50m), 0.65% (£50m-£250m), and 0.55% (£250m+). Based on net assets of approximately £810 million at period-end, the blended effective fee rate is approximately 0.567%. Compared with the prior-year period, when net assets were approximately £614 million and the blended fee rate was approximately 0.60%, the fee rate declined as scale increased, reflecting the economies of scale inherent in the tiered design. However, the total half-year management fee is not directly disclosed and must be estimated from the midpoint of opening and closing net assets: approximately (6.13+8.10)/2 = 7.12 (in £100m) × 0.567% / 2 ≈ £2.02 million, which is less than 1% of net returns (£217m), indicating reasonable fee control.
2. Capital Nature of the Tax Burden
In the current period, overseas withholding tax of £305,000 (prior-year: £428,000) was booked as income tax, but the larger tax expense was India capital gains tax of £3.535 million (paid in cash), corresponding to the "capital tax items" in Note 5 as withholding related to Indian transactions. Notably, India tax in the prior-year period was only £198,000; the sharp increase this period suggests either changes in Indian market regulation or tax rules, or an acceleration in the company's disposal of its India positions (consistent with the large-scale buying and selling in investing activities). This tax item directly reduces capital returns. If the company continues to adjust its India holdings, future tax costs may remain elevated.
3. Warning from Negative Income Return
The income statement shows a current-period income return loss of £1.025 million (-1.22p per share), while capital returns were as high as £218 million. This continues the company's profile of "capital appreciation primary, income secondary," but the loss has widened (prior-year period income was +£467,000). Reasons include:
The company cut its final dividend from 2.65p to 1.50p and paid no interim dividend, a direct reflection of the pressure on the income side. For traditional trust investors who rely on dividends, this strategy increasingly resembles a "growth" fund.
Note 8 describes the three levels of fair value, but does not provide a specific breakdown of amounts by level (the subsequent table is not presented). Based on the company's prior disclosures (2025 annual report), the vast majority of investments are listed equities in Japan and Asia-Pacific, classified as Level 1 (quoted prices in active markets). As of January 31, 2026, the portfolio contained holding gains of as much as £299 million (included in capital reserves), compared with £112 million in the same period of the prior year, reflecting a substantial rise in market value. However, this also implies a higher proportion of unrealized gains — if the market reverses, the drawdown in capital reserves could be more pronounced. Attention should be paid to the amount distribution by level in the subsequent notes to confirm whether any Level 3 (unlisted or liquidity-restricted) assets exist.
In its going concern statement, the board specifically highlighted "heightened market volatility due to macroeconomic and geopolitical concerns," but concluded that these do not affect the company's ability to continue as a going concern. The assets consist mainly of listed securities that can be readily liquidated, and liabilities remain far below assets. However, a key juncture is the continuation vote in November 2026, which takes place every five years. If shareholders vote against it at that time, the company could face liquidation or reform. Strong half-year results (NAV up 36%) may help the board win back support, but persistently weak income and rising leverage could give opponents grounds.
In addition, all borrowings require board approval, and gearing levels are reviewed regularly. At the period end, the net debt/net assets ratio was not listed in the statements, but based on the borrowing balance (opening +24,220) and net assets of 809,883 ten-thousand pounds, net borrowing is estimated at about 6,000 ten-thousand+? In fact, the cash flow statement also shows that the opening borrowing balance existed but was not directly given; however, inferred from interest expense, average borrowing is about 1,200-1,300 ten-thousand (assuming an interest rate of 5%). More complete notes are needed to confirm the leverage multiple. That said, the difference between total assets and net assets disclosed in the 2025 annual report would reveal total borrowings; this cannot be accurately inferred here.
| Metric | 2026 H1 | 2025 H1 | 2025 FY (audited) |
|---|---|---|---|
| Income return (£'000) | (1,025) | 467 | 2,090 |
| Capital return (£'000) | 218,485 | 17,160 | 41,461 |
| Total net return (£'000) | 217,460 | 17,627 | 43,551 |
| Income per share (p) | (1.22) | 0.52 | 2.34 |
| Capital per share (p) | 260.43 | 18.97 | 46.42 |
| Total return per share (p) | 259.21 | 19.49 | 48.76 |
| Weighted average shares (millions) | 83.89 | 90.44 | 89.32 |
The half-year total return is already 5 times that of the full FY2025 (43,551), clearly benefiting from the strength of Japanese and Asia-Pacific technology stocks (e.g., semiconductors, AI supply chain, etc.). However, the capital return per share of 260.43p far exceeds the NAV increase (976.21 vs 716.13 = +260.08p), almost entirely contributed by capital gains. Reviewing the opening NAV and capital gains, one can infer a total portfolio return of roughly 218,485 / (613,357+20,367+capital?) ≈ 35% or so, consistent with the description in the report's investment section (if mentioned).
Going forward, attention should be paid to the borrowing balance, the effective leverage ratio, and the investment breakdown by fair-value hierarchy in the notes, in order to fully assess risk exposure.
Revolving credit borrowings rose from £35.78M at the start of the period to £60.0M, reaching the cap on the one-year multi-currency facility — an increase of 67.7%. This change closely matches the share buyback spend over the same period (£19.674M): approximately 80% of the new borrowings went to buybacks rather than to adding portfolio positions. This conveys a strategic signal worth noting: management chose to support buybacks with credit rather than by selling holdings, indicating an unwillingness to cut equity positions when valuations may be low, thereby maintaining full exposure to Asian markets while still delivering cash returns to shareholders.
The cost of this operating model is higher financial risk. The table below compares the key debt and buyback data at the start and end of the period:
| Item | 31 January 2026 | 31 July 2025 | Change |
|---|---|---|---|
| Revolving credit borrowings | £60,000,000 | £35,780,000 | +67.7% |
| Credit facility | £60,000,000 | £60,000,000 | Fully drawn |
| Half-year buyback spend | £19,674,000 | £29,766,000 (full year) | Six months at 66% of full year |
| Debt/total portfolio value | 6.9% | 5.5% | +1.4pp |
Another date that deserves close attention is March 2026 — the credit facility matures only a few weeks after the reporting date. The company did not disclose a renewal arrangement in the report, which means management faces three options: extend it, repay it from portfolio cash flows, or passively reduce positions amid market volatility. Given that investment trusts generally maintain permanent capital structures, short-term debt is more a tactical tool than a long-term reliance. The real risk is that maturity coincides with a sharp market correction, in which case "selling shares cheaply to repay debt" and "buying back at a discount to a low NAV" would create an awkward lose-lose situation.
The tax provision fell from £4.191M to nil, with the disclosed reason being that there were no net unrealized taxable capital gains. This is an information-rich yet easily overlooked data point — it indicates that the unrealized gains in the Indian portfolio disappeared completely within six months. Possible causes include: a decline in the market value of holdings, sales of profitable positions, or the offsetting of existing gains through stop-losses.
| Item | 31 January 2026 | 31 July 2025 |
|---|---|---|
| India capital gains tax provision | nil | £4,191,000 |
| Applicable tax rates (long-term/short-term) | 12.5% / 20% | 12.5% / 20% |
| Implied unrealized taxable gains | approximately £0 | approximately £33.5M (calculated at 12.5%) |
This "zeroing" is positive on the tax front — the company temporarily avoids a potential future tax liability. On the investment front, however, it implies that the Indian exposure experienced a notable price drawdown or position restructuring during the period. Given that trading costs also rose sharply in the report, the Indian market positions may have undergone a concentrated "clean-out" adjustment.
What deserves attention going forward: if the Indian market rebounds in the next phase, the provision will be recognized again, and its size will in turn serve as a "thermometer" for the recovery of profitability in Indian positions. In addition, the company distinguishes between long-term (12.5%) and short-term (20%) capital gains tax rates for its India investments, which will affect the timing preference for future sales — if holdings are held for a full year before selling, the tax burden falls significantly. This may objectively incline the company toward holding Indian positions for longer.
The details embedded in the share capital movements are richer than the surface numbers suggest. The half-year buyback of 2,686,152 shares represented 3.1% of the issued share capital at the start of the period; on an annualized basis this is approximately 6.2%, higher than the prior year's 5.5% — the intensity of buybacks is increasing. More important, however, is the jump in the average buyback price:
| Period | Shares repurchased | Buyback cost | Average price (£/share) |
|---|---|---|---|
| August 2025 – January 2026 | 2,686,152 | £19,674,000 | £7.32 |
| August 2024 – July 2025 | 5,008,324 | £29,766,000 | £5.94 |
| Average price ratio | — | — | +23.2% |
The 23.2% rise in the average buyback price reflects both NAV growth and the fact that, even after a substantial share price increase, the company still viewed the shares as trading at a sufficient discount to net asset value to justify continued buybacks. This is an important confidence signal — the buyback did not stop because the share price rose; instead, it maintained a faster pace of execution at higher prices.
Treasury shares increased from 6.43M shares to 9.11M shares, while total share capital remained unchanged at 92.07M shares over the same period, indicating that the company is buying back and parking shares in treasury while retaining "ammunition" for future premium issuance. This can be viewed as a capital management strategy: buy and hold at a discount, release at a premium, and thereby create NAV-per-share accretion for existing shareholders at every stage. Together with the remaining buyback authorization of 12.34M shares (approximately 14.9% of current issued share capital), the company still has ample room for capital operations.
After the period end, an additional 236,901 shares were repurchased through 10 March, indicating that this strategy is still being executed and has not yet turned.
The change in trading costs deserves to be examined separately. Half-year purchase transaction costs reached £243,000, 11.6 times the level of the same period last year; sales transaction costs were £275,000, 9.2 times the prior-year level. Total transaction costs of £518,000 have already exceeded the full-year total of the prior year, £452,000 (190+262).
| Item | 31 January 2026 (six months) | 31 January 2025 (six months) | 31 July 2025 (full year) |
|---|---|---|---|
| Purchase transaction costs | £243,000 | £21,000 | £190,000 |
| Sales transaction costs | £275,000 | £30,000 | £262,000 |
| Total | £518,000 | £51,000 | £452,000 |
| Total/period-end portfolio | 0.06% | ~0.01% | 0.07% |
The drivers of the sharp increase in trading costs may not be a single rebalancing move but a combination of factors: adjustment of Indian positions, building positions in new market opportunities, and potential portfolio rebalancing. Corroborating evidence is that total portfolio value rose sharply from £651.8M to £866.6M, an increase of 33% — even after stripping out the contribution of market appreciation, a change of this magnitude implies that the internal structure of the portfolio underwent significant reconstruction.
Trading volume increased, but the cost ratio remained contained at 0.06%, indicating that execution efficiency has not deteriorated. However, if the same turnover intensity is maintained in the second half, full-year trading costs will exceed £1M. For a trust whose total expense ratio is already low, this could become a new variable that erodes net performance.
Taken together, this period's financial data paints a management picture of "aggressive buybacks + debt support + portfolio reconstruction." But all of these operations point to a question that must be answered: the £60M credit facility matures in March 2026 — how does the company intend to respond?
From a cash-flow perspective, the company holds a large amount of listed securities (£826.8M) and has ample liquidity; it could certainly repay the debt by selling some positions. But if that were the intention, why use borrowed money to buy back shares rather than directly trimming positions? The most reasonable explanation is that management wants to defer the "sell decision" until market conditions are more favorable, while using the short-term credit window to capture the buyback opportunity presented by the current discount in the share price. This is a classic "time arbitrage" — using inexpensive short-term debt in exchange for a longer window of portfolio optimization.
Whether this strategy succeeds depends on two variables in the coming months: the direction of Asia-Pacific and Indian markets, and the availability of a credit extension. If markets recover as management expects, a rebound in Indian positions would simultaneously resolve the provision issue and provide a source of repayment. If markets remain weak, the company could face the passive situation of "being chased to repay debt right after buying back at a discount." Although the financial notes in this report do not provide a direct answer, the leverage and tax signals between the lines have already laid the groundwork for the next phase.
The continuation defines 11 terms/APMs in total, a density far higher than that of a typical listed company report. This reflects a distinctive trend in the investment trust industry: because NAV per share is the most intuitive performance yardstick for investors, trust management has an obligation to provide itemized disclosure of the composition of NAV, its adjustment methods, and derived indicators. Three structural features are worth noting:
| Feature | Specific manifestation | Information increment |
|---|---|---|
| Explicit definitional basis | Total assets defined as "total assets less current liabilities (excluding borrowings)" | Separates "total assets" from the "NAV calculation basis," avoiding the misleading effect of leverage positions |
| Transparent calculation process | Discount/Premium calculation embedded directly in the table | Investors can verify the calculation themselves, enhancing data credibility |
| Extreme cases covered | "In periods where no dividend is paid, total return equates to capital return" | Removes comparability barriers across different dividend policies |
The APM table provides two sets of data — 31 January 2026 and 31 July 2025 (audited) — while the Total Return table provides four sets (two time points × NAV/Share Price dimensions). This multi-basis disclosure strategy implies:
This report simultaneously discloses three metrics — Net Gearing (7%), Gross Gearing (7%), and Leverage (Gross method) — a practice rarely seen among small- and mid-cap investment trusts, reflecting BG's high standards of compliance with UK AIFMD. The data show:
| Metric | 31 Jan 2026 | 31 Jul 2025 | Change |
|---|---|---|---|
| Carrying value of borrowings | £60.0m | £35.8m | +67.7% |
| Net gearing | 7% | 5% | +2pct |
| Gross gearing | 7% | 6% | +1pct |
| Cash and cash equivalents | £6.9m | £3.8m | +80.7% |
The increase in borrowings and the increase in cash are almost synchronized, indicating that part of the new borrowings was retained as a liquidity buffer rather than fully deployed into the market. This reveals management's precautionary preparation for downside risk when NAV is at elevated levels — subtly echoing the judgment, possibly stated in the first half of the report, that "technology valuations are high but structural opportunities remain."
The practice of deducting "sales for subsequent settlement" from borrowings while adding back "purchases for subsequent settlement" effectively incorporates the net amount of unsettled trades into the leverage calculation. As of 31 Jan 2026, purchases awaiting settlement of £4.7m were significantly larger than sales awaiting settlement of £0, indicating a net buying position at the fiscal year-end — further confirming the conclusion that "positions were still being added at the end of the reporting period."
The appearance of active share as an APM in the Interim Report is itself an externalization of BG's investment philosophy. Most investment trusts, by convention, disclose only performance and fees; BG deliberately introduced this metric, conveying a dual signal:
1. Strategic orthodoxy: A high active share is the quantitative basis for the "reason for being" of active management. In a UK market where passive investing is increasingly prevalent, this is an implicit answer to "why charge higher management fees."
2. Self-positioning relative to the benchmark: Pacific Horizon's comparison benchmark is typically the MSCI All Country Asia Pacific Index. If active share remains high over the long term (BG's trusts are typically above 90%), it indicates that this is essentially a "benchmark-deviation" product — investors' expectation management should be based on alpha generation rather than beta tracking.
If the Interim Report is viewed as a complete communication text, the placement of the Glossary — after the disclaimer and before the company information — has functional significance:
```
Investment narrative (strategy, holdings, performance)
↓
Risk isolation (disclaimer)
↓
Term standardization (APM definitions)
↓
Corporate governance (directors, auditor, depositary)
```
This sequence guides readers through four progressive layers of information — "Why invest" → "What if risk" → "How to read" → "Who governs." Compared with a direct listing of legal elements, this arrangement is more conducive to retail investors' understanding of structural risks.
A data contradiction worth pointing out, implicit in the table, is that the period-end dividend fell from 2.65p in the same period last year to 1.50p (a decline of 43.4%), while NAV total return over the same period was as high as 36.6%. Combined with the company's simultaneous conversion of some shares into Treasury shares through buybacks (which boosts earnings per share), one can infer that the board judged the market price's discount to NAV (8.1%) to offer investors a better path to returns through capital returns rather than dividends — not so much a "stingy dividend policy" as a capital allocation strategy based on the discount.
Part 6/6 closes the document with three blocks of "non-narrative" content (disclaimer, glossary, company information). Seemingly plain, they elevate this interim report from mere "performance disclosure" into a "governance document" with traceability, verifiability, and legal completeness.
From the Introduction to the Company Information, the entire report completes the following loop:
| Narrative stage | Text vehicle | Function |
|---|---|---|
| Proposition (What we did) | Investment commentary in the first half | Builds trust and expectations |
| Measurement (What we achieved) | Financial Statements + APM | Provides verifiable quantitative evidence |
| Boundaries (What we are not liable for) | Disclaimer | Defines the scope of liability |
| Tools (How to interpret) | Glossary + APM definitions | Unifies the investor cognitive framework |
| Accountability (Who is accountable) | Directors & Auditor | Establishes accountability |
The design logic behind this structure is that the long-term competitiveness of an actively managed investment trust depends not only on performance alpha but also on "disclosure alpha" — that is, a governance premium that surpasses peers in transparency and in the explanatory framework. The 36.6% NAV growth that Pacific Horizon delivered in the interim period of fiscal 2026 is precisely a result built on this sophisticated disclosure system, and its credibility therefore far exceeds that of a set of simple numbers.